跳转至

LinkedIn Profile 指标追踪

状态: 🚀 进行中

创建日期: 2026-02-20 最后更新: 2026-03-27 (session 29) — 发布 TPM 判断力 vs AI 数据帖(含配图)并完成 X standalone cross-post + Reply(LinkedIn URL)


📅 日历事件

事件名称 开始 结束 地点 日历 提醒 备注
PMI PMXPO 2026 (Virtual) 2026-03-26 07:00 2026-03-26 11:45 Online -60 pmi.org/events/pmxpo;免费;PM 全球最大虚拟活动;可获 PDU
ProductCamp Vancouver 2026 2026-05-09 2026-05-09 Vancouver, BC -1440 productcampvancouver.org;免费 unconference;PM/TPM networking
Vancouver AI Meetup (June) 2026-06-25 2026-06-25 TBD, Vancouver, BC 📄 - 需提前订阅 lu.ma BC+AI Events 抢票;March/April 均售罄

日历状态说明:✅=已加入 / 📄=仅文档


追踪目标

通过 profile 优化(重写 Headline/About/Experience、重建 Skills、定期发帖),将 LinkedIn 搜索曝光和 Profile 访问量提升 10x。


基准快照(2026-02-20 优化前)

指标 数值 备注
Profile views / 7 days 5 2026-02-20 读取
Search appearances / 7 days 2 2026-02-20 读取
Post impressions / 7 days 1 2026-02-20 读取
Connections 326 2026-02-20 读取
Followers 356 2026-02-20 读取

目标(优化后 4 周)

指标 4周目标 3个月目标
Profile views / 7 days 40+ 100+
Search appearances / 7 days 20+ 50+
Post impressions / 帖 500+ 2000+
Followers 400+ 600+

里程碑记录

日期 里程碑 状态
2026-02-20 Headline 更新为 Option A($30.8B TPM) ✅ 完成
2026-02-20 About 完全重写(TPM定位,数字密集) ✅ 完成
2026-02-20 Alibaba Experience 拆分为 3 段 ✅ 完成
2026-02-20 IELTS Test Score 删除 ✅ 完成
2026-02-20 Open to work 更新为 Senior TPM ✅ 完成
2026-02-20 Itsail Studio 描述精简 ✅ 完成
2026-02-21 Skills 置顶 Program Management / 删除 Senior PM ✅ 完成
2026-02-20 Itsail Studio 就业类型 Co-op → Self-employed ✅ 完成
2026-02-20 RealMaster 描述加入 PM 角度 ✅ 完成
2026-02-21 发布第一篇英文帖(TPM war story · 1111 Day + AI反思) ✅ 完成
2026-02-21 发送同行 TPM 连接请求 5 人(批次一:Amazon/Google/Microsoft) ✅ 完成
2026-02-21 发送阿里背景连接请求 8 人(批次二:搜索结果 2 页) ✅ 完成
2026-02-22 发送 Vancouver/Seattle TPM 连接请求 6 人(批次三) ✅ 完成
2026-02-22 发送 Vancouver/Seattle 扩展连接请求 7 人(批次四) ✅ 完成
2026-02-23 发送 Seattle/Vancouver 扩展连接请求 8 人(批次五) ✅ 完成
2026-02-23 发送 Seattle ex-Alibaba/TPM 连接请求 8 人(批次六) ✅ 完成
2026-02-25 发送 Vancouver/Seattle Sr TPM 连接请求 8 人(批次七) ✅ 完成
2026-02-27 发送 Vancouver/Seattle/Canada Sr TPM 连接请求 8 人(批次八,全部成功,累计 58) ✅ 完成
2026-02-27 发送 Vancouver/Seattle/US Sr TPM+谷歌/微软/字节 TPM 连接 10 人(批次九,全部成功,累计 68) ✅ 完成
2026-02-27 发送 Seattle/US Google/Microsoft/TikTok Sr TPM 连接 10 人(批次十,全部成功,累计 78) ✅ 完成
2026-02-27 发送 Amazon Principal + Meta TPM 连接 10 人(批次十一,全部成功,累计 88) ✅ 完成
2026-02-27 发送 Vancouver 本地科技公司 TPM 连接 10 人(批次十二,全部成功,累计 98) ✅ 完成
2026-02-27 发送 Amazon Principal + Apple + Expedia TPM 连接 10 人(批次十三,全部成功,累计 108) ✅ 完成
2026-02-22 Banner 图片更新(Canva) ✅ 完成
2026-02-28 发送 ex-Alibaba 开发者连接 7 人(批次十四,全部成功,累计 115) ✅ 完成
2026-02-28 发送 ex-Alibaba 开发者连接 5 人(批次十五,全部成功,累计 120) ✅ 完成
2026-02-28 发送 HDU 校友连接 5 人(批次十六,全部成功,累计 125) ✅ 完成
2026-02-28 发送 HDU 校友连接 3 人(批次十七,全部成功,累计 128) ✅ 完成
2026-02-28 发布 5 条 career 帖子评论(去 AI 味,10-30 词) ✅ 完成
2026-03-01 LinkedIn → X thread 发布(3条,去AI味儿) ✅ 完成
2026-03-01 X 上发布 5 条 TPM/career 帖子互动评论 ✅ 完成
2026-03-01 发送连接请求 5 人(session 12) ✅ 完成
2026-03-02 发送 ex-Alibaba 开发者连接请求 1 人(Kris Yang,pending) ✅ 完成
2026-03-03 发布 LinkedIn 帖子(Claude 宕机 → AI 单点故障) ✅ 完成
2026-03-03 LinkedIn → X thread cross-post(3条) ✅ 完成
2026-03-04 发送 ByteDance TPM 连接请求 14 人(批次十九前半) ✅ 完成
2026-03-04 发送 Amazon/Vancouver/Seattle TPM 连接请求 6 人(批次十九后半) ✅ 完成
2026-03-04 发布 LinkedIn 帖子(GPT-5.4 → 规划地平线缩短) ✅ 完成
2026-03-04 LinkedIn → X thread cross-post(3条) ✅ 完成
2026-03-11 发布评论:Aaron Hodes logistics/opportunity cost 帖(11.11 Alibaba PM hours → carrier escalations 换算) ✅ 完成
2026-03-11 发布 LinkedIn 帖子(AgenticAI × TPM 协调问题)+ X standalone post + Reply 放 LinkedIn URL ✅ 完成
2026-03-12 发布评论:Jeff Oriecuia(AWS Sr. TPM, Vancouver)AI live demo 帖("Live demos always find their edge cases…") ✅ 完成
2026-03-14 发布评论:Aleksandr Stepanov(Claude Code dev platform帖)"At some point I stopped being the one who gives instructions..." ✅ 完成
2026-03-16 发布 LinkedIn 帖子(TPM 3问题框架 × AI Agents:430工程师 / 可见性 / 依赖地图)+ X standalone post + Reply 放 LinkedIn URL ✅ 完成
2026-03-16 发布评论:Aiishvar Chandra(PM最大误解帖,119 comments)"Most people think PM is about coordination. It isn't..." ✅ 完成
2026-03-16 发布评论:Chris Stasiuk(Junior engineer frustrated帖,8 comments)"At 430 engineers, I kept building better status dashboards..." ✅ 完成
2026-03-18 发布评论:Priya Raman(FDE热门帖,19 comments)"This is what happened to TPM too. Once coordination ran through agents..." ✅ 完成
2026-03-19 发布 LinkedIn 帖子(16.5s→5.4s 手淘性能优化 × AI skills 热点反转)+ X standalone post + Reply 放 LinkedIn URL ✅ 完成
2026-03-23 发布 LinkedIn 帖子(AI 对 IT 各工种冲击,8 roles infographic)+ X standalone post + Reply 放 LinkedIn URL ✅ 完成
2026-03-23 发布评论:Alex Xu(ByteByteGo,Load Balancer vs API Gateway帖,723 reactions)"We had someone put auth logic in the load balancer once..." ✅ 完成
2026-03-23 发布评论:Nick Palasz(Slyleadz,77 Cold Email Openers帖,274 reactions)"AI writes most cold outreach now..." ✅ 完成
2026-03-26 发布评论:Nilesh Naik(AI-First TPM Leader,effort estimation with Claude Skills,23 reactions)"We built something similar for sprint scoping across 109 services..." ✅ 完成
2026-03-26 发布评论:Arpit Shah(TPM at Google,7 AI Prompts for T/PM,104 reactions)"Prompt #1 (priority re-ranking) is where I got the most value..." ✅ 完成
2026-03-27 发布 LinkedIn 帖子(TPM 判断力 vs AI 数据)+ X standalone post + Reply 放 LinkedIn URL ✅ 完成

周度指标追踪

每周五记录一次,对比前一周变化。

记录日期 Profile Views/7d Search App./7d Post Impressions/7d Members Reached Followers Connections 备注
2026-02-20 5 2 1 356 326 基准(优化启动)
2026-02-21 —¹ 11 ↑450% 56 ↑5,500% 9 356 326 优化+首帖发布后第1天
2026-02-22 14 ↑180% 107 ↑11,600% 365 ↑2.5% 335 ↑2.8% session 5 复测(批次三+四执行后)
2026-02-23 21 ↑320% 135 ↑13,400% 370 ↑3.9% 340 ↑4.3% session 7(批次五+六执行后)
2026-02-25 26 ↑420% 155 ↑15,400% 377 ↑5.9% 347 ↑6.4% session 8(批次七执行后)
2026-02-27 33 ↑560% 179 ↑17,800% 380 ↑6.7% 350 ↑7.4% session 9(批次八执行后,8/8 全部成功)
2026-02-27 45 ↑800% 5 ↑150% 252 ↑25,100% 67 389 ↑9.3% 358 ↑9.8% session 10(批次十一/十二/十三执行后,累计108个请求;帖子 Impressions 含 AI替代TPM 新帖)
2026-02-27 59(90天口径) 5 ↑150% 234 ↑23,300% 67 397 ↑11.5% 367 ↑12.6% session 11 日终复测;Profile appearances 677/week;Followers +8、Connections +9 vs session 10
2026-02-28 69(90d口径) 287 ↑154% 105 ↑228% 399 ↑11.9% 396 ↑21.5% session 11(批次十四/十五/十六/十七执行后,累计128;5条帖子互动)
2026-03-01 90(90d口径) 485 412 ↑3.3% 382 session 12(X thread发布;5条X互动;连接请求5人)
2026-03-03 93(90d口径) 5/7d(677 all appearances) 638 ↑31.6% 273 ↑534.9% 419 ↑1.7% 389 ↑1.8% session 14(Claude宕机帖;X thread;Rose互动;Search App 0%增长—需关注)
2026-03-04 97(90d口径) 7/7d(823 all appearances ↑22%) 771 ↑333.2% 423 ↑1.0% 420 ↑8.0% session 15(GPT-5.4帖;X thread;批次十九 20人 ByteDance+Amazon)
2026-03-05 114(90d口径) 7/7d(823 all appearances) 1,230 ↑59.5% 433 ↑2.4% 430 ↑2.4% session 16(AI无限产能帖;X thread;全量指标更新;Post impressions 7d ↑331.6%)
2026-03-08 120(90d口径) 7/7d(823 all appearances;●0%,数据区间2/24-3/3) 1,345 ↑9.3% 436 ↑0.7% 433 ↑0.7% session 17(数据口径确认;Jay Tze 回复 WeChat;URL修正 xin-peng-tpm;连接管理页与Profile页差值28已记录说明)
2026-03-09 121(90d口径) 7/7d(823 all appearances;●0%,页面仍显示上一完整周) 1,337 ↓0.6% 439 ↑0.7% 436 ↑0.7% session 18(live profile review后复测;Taobao title 已修正;Top skills 已调整但展示卡片可能有缓存)
2026-03-11 123(90d口径) 23/7d(1,537 all appearances ↑87%;数据区间3/3-3/10) 1,044 ↓21.9% 441 ↑0.5% 438 ↑0.5% session 19(Search 23 ↑228% vs 上周7;Profile appearances 1,537大幅提升;Post impressions下降因2/27高峰帖滑出7d窗口;Aaron Hodes logistics帖评论发布)
2026-03-12 126(90d口径) 23/7d(1,537 all appearances;0%,数据区间未更新) 380 ↓63.6% 444 ↑0.7% 440 ↑0.5% session 20(AgenticAI帖发布后第1天;Post impressions低因新帖刚发;Jeff Oriecuia live demo帖评论发布)
2026-03-16 127(90d口径) 23/7d(1,537 all appearances;●0%,数据区间3/3-3/10 未更新) 175 ↓87% 449 ↑1.1% 445 ↑1.1% session 22(草稿四发布:TPM 3问题×AI Agents;X standalone post;Post impressions低因新帖发布当天正常)
2026-03-18 129(90d口径) 9/7d(1,138 all appearances ↓25%;数据区间3/10-3/17) 149 ↓15% 449 ●0% 448 ↑0.7% session 23(数据区间更新到3/10-3/17;Search App 23→9 ↓61%;All appearances 1,537→1,138 ↓26%;Priya Raman FDE帖评论发布)
2026-03-19 129(90d口径)●0% 9/7d(1,138 all appearances;●0%,数据区间3/10-3/17 未更新) 153 ↑2.7% 452 ↑0.7% 448 ●0% session 24(发布新帖:16.5s→5.4s 手淘性能优化×AI skills热点;X cross-post + Reply放LinkedIn URL;历史帖子Impressions补录)
2026-03-20 129(90d口径)●0% 9/7d(1,138 all appearances;●0%,数据区间3/10-3/17 未更新) 183 ↑19.6% 452 ●0% 448 ●0% session 25(全量指标更新;Post impressions 7d 153→183 ↑19.6%;优化后第4周)
2026-03-21 129(90d口径)●0% 9/7d(1,138 all appearances;●0%,数据区间3/10-3/17 未更新) 201 ↑9.8% 453 ↑0.2% 449 ↑0.2% session 26(全量指标更新;Post impressions 7d 183→201 ↑9.8%;3/19新帖 11→34 +209%;优化后第4周+1天)
2026-03-23 129(90d口径)●0% 9/7d(1,138 all appearances;●0%,数据区间3/10-3/17 未更新) 568 ↑182.6% 454 ↑0.2% 450 ↑0.2% session 27a(全量指标更新;Post impressions 7d 201→568 ↑182.6%;AI各工种新帖发布后~1h即192 impressions;2条评论发布;LinkedIn playbook 新增评论禁止简历式开头规则)
2026-03-23 130(90d口径)↑0.8% 9/7d(1,138 all appearances;●0%,数据区间3/10-3/17 未更新) 626 ↑10.2% 455 ↑0.2% 450 ●0% session 27b(+2h复测;Post impressions 7d 568→626 ↑10.2%;Profile viewers 129→130 首次变化;AI各工种帖 192→239 ↑24.5%;3/19帖 34→211 ↑520.6%;3/16帖 96→129 ↑34.4%;3/11帖 46→74 ↑60.9%)
2026-03-26 130(90d口径)●0% 9/7d(1,138 all appearances;●0%,数据区间3/10-3/17 未更新) 696 ↑11.2% 309 ↑263.6% 455 ●0% 450 ●0% session 28(全量指标更新;Post impressions 7d cumulative 696 ↑349.1% vs prior 7d;日趋势:Mar23=534 → Mar24=609 → Mar25=683 → Mar26=696;AI各工种帖 306 imp, 2 reactions;3/19帖 243 imp;2条新评论:Nilesh Naik TPM effort estimation + Arpit Shah Google TPM 7 prompts)

¹ Dashboard 于 2026-02-21 起将 Profile Views 口径改为 Past 90 days,无法与基准 7d 数据对比,故填

取数口径(固定,不得随意更改): - Followers / Post impressions / Profile viewerslinkedin.com/dashboard/ → Analytics 区块 - Connectionslinkedin.com/mynetwork/invite-connect/connections/ 管理页顶部(实时值) - Search appearances / All appearanceslinkedin.com/analytics/search-appearances/(显示上一完整周,非实时) - ⚠️ 禁止用 Profile 主页(/in/xin-peng-tpm/)显示的 connections 数字——该值为缓存,通常滞后且偏低(例:2026-03-07 管理页=433,Profile页=405,差值28)


帖子 Analytics 追踪

发布日期 类型 主题 Impressions Members Reached Reactions Comments Reposts 备注
发布日期 类型 主题 Impressions Reactions Comments Reposts 备注
--------- ------ ------ ------------- ----------- ---------- --------- ------
2026-02-21 Post 1111 Day war room + AI 反思 120 0 0 0 urn🇱🇮activity:7430771034049486848;session 26 更新(119→120)
2026-02-23 Post OpenAI 对抗审稿帖 77 1 0 0 urn🇱🇮activity:7431551382945673216;session 26 更新(76→77)
2026-02-25 Post 跨团队依赖盲区(Vancouver/Seattle TPM CTA) 97 0 0 0 urn🇱🇮activity:7432524877825462273;session 26 更新(96→97)
2026-02-27 Post AI 会替代 TPM 吗?(humanizer 优化版) 810 2 2 0 urn🇱🇮activity:7433234702435770368;最高表现帖;Rose Ruoxi Liu + Shilpi Midha 点赞;Rose 评论"This couldn't resonate with me more!";session 26 更新(809→810)
2026-02-28 Post Vibe coding / AI dependency risk($30B + intern) 167 2 0 0 urn🇱🇮activity:7433371372258721792;session 26 更新(165→167)
2026-03-02 Post AI 替代了我讨厌的工作 / 2017 Double 11 判断力帖 230 0 0 0 urn🇱🇮activity:7434353485120282624;session 26 更新(229→230)第二高表现帖
2026-03-03 Post Claude 宕机 → TPM 单点故障 / AI 依赖风险 97 0 0 0 urn🇱🇮activity:7434706292834783232;session 26 更新(96→97)
2026-03-04 Post GPT-5.4 → 规划地平线缩短 / 12月roadmap vs 90天迭代 153 0 0 0 urn🇱🇮activity:7435055795161124864;session 26 更新(152→153)
2026-03-05 Post AI 无限产能幻觉 → 效率工具上瘾 / 个人反思 102 1 0 0 urn🇱🇮activity:7435511590675951616;session 26 更新(101→102)
2026-03-08 Post AI Engineer 是最大风险 / 传统 PM 直觉 vs AI 项目管理差异 192 2 2 0 urn🇱🇮share:7436471250195107841;session 26 更新(190→192);#TPM #AIProjects;X thread同步(3条)
2026-03-11 Comment Aaron Hodes logistics帖:11.11 PM hours被carrier escalations吞噬的数字化换算 帖子地址:https://www.linkedin.com/posts/aaronhodes_your-ceos-biggest-logistics-concern-is-probably-share-7437550756360253441-B1_p
2026-03-11 Post AgenticAI × TPM 协调问题:43 sprints / 430 engineers / week 6 API 冲突 / Singles' Day 74 4 0 0 urn🇱🇮activity:7437703504753807360;X cross-post: https://x.com/XINDR365/status/2031938504188703060;session 27b 更新(46→74,↑60.9%)
2026-03-12 Comment Jeff Oriecuia(AWS Sr. TPM, Vancouver)AI live demo帖 session 20
2026-03-14 Comment Aleksandr Stepanov(Claude Code dev platform帖) session 21
2026-03-16 Post TPM 3问题框架 × AI Agents:430工程师 / 可见性 / 依赖地图 129 0 0 0 urn🇱🇮activity:7439332772328095745;X cross-post;session 27b 更新(96→129,↑34.4%)
2026-03-16 Comment Aiishvar Chandra(PM最大误解帖,119 comments) session 22
2026-03-16 Comment Chris Stasiuk(Junior engineer frustrated帖) session 22
2026-03-18 Comment Priya Raman(FDE热门帖,19 comments) session 23
2026-03-19 Post 16.5s→5.4s 手淘性能优化 × AI skills热点反转 / 26团队政治 243 0 0 0 urn🇱🇮activity:7440598965290610688;X cross-post: https://x.com/XINDR365/status/2034833639226450334;#TPM #ProgramManagement #AISkills;session 28 更新(211→243,Day 7)
2026-03-23 Post AI 对 IT 各工种冲击:8 roles / QA script writers / DBA 2am / TPM judgment 306 2 0 0 urn🇱🇮activity:7441930902932406272;X cross-post: https://x.com/XINDR365/status/2036166937693265968;配图:linkedin_post_ai_impact_roles_20260323.png;session 28 更新(239→306,发布后 Day 3,2 reactions)
2026-03-23 Comment Alex Xu(ByteByteGo)Load Balancer vs API Gateway帖 urn🇱🇮activity:7441872725410824192;session 27
2026-03-23 Comment Nick Palasz(Slyleadz)77 Cold Email Openers帖 urn🇱🇮activity:7440297947478765568;session 27
2026-03-26 Comment Nilesh Naik(AI-First TPM Leader)effort estimation with Claude Skills帖,23 reactions session 28
2026-03-26 Comment Arpit Shah(TPM at Google)7 AI Prompts for T/PM帖,104 reactions session 28
2026-03-27 Post TPM 判断力 vs AI 数据:green status + midnight judgment + payment escalation 0 0 0 0 urn🇱🇮share:7443332174352936960;X cross-post: https://x.com/XINDR365/status/2037567448686481411;配图:linkedin_post_tpm_judgment_vs_ai_20260327.png;session 29

Followers Demographics 分析与改进优先级

快照对比(2/25 → 3/3 → 3/24)

维度 2026-02-25 2026-03-03 2026-03-24 20天变化 (3/3→3/24)
Job Title #1 Search Consultant 12.3% Search Consultant 11.2% Search Consultant 10.1% -1.1%(持续下降)
Job Title #2 HR Specialist 2.4% TPM 4.9% TPM 7.7% ⬆️ +2.8%(+57%)
Job Title #3 Software Engineer 2.9% Software Engineer 4% +1.1%
Job Title #4 HR Specialist 2.2% HR Specialist 2% -0.2%
Job Title #5 Founder 1.9% Recruitment Specialist 1.8% 新进入
Location #1 杭州 23.2% 杭州 21.2% 杭州 19.4% -1.8%(持续下降)
Location #2 深圳 7.1% Greater Seattle 8.6% ⬆️ Seattle 升至 #2(+2.0%,+30%)
Location #3 Greater Seattle 6.6% 深圳 6.6% ●0%
Location #4 朝阳 5.8% Greater Vancouver 6.2% ⬆️ Vancouver 升至 #4(+0.8%,+15%)
Location #5 Greater Seattle 3.7% Greater Vancouver 5.4% 朝阳 5.1% 朝阳跌出 #4
Industry: HR 21.6% 19.7% 17.8% -1.9%(持续下降)
Industry: Staffing 10.4% 9.5% 8.6% -0.9%
Industry: Software Dev 7.5% 10% 11.2% +1.2%
Amazon(公司) 1.9% 3.4% 4.4% ⬆️ +1.0%(+29%)
ByteDance 1.9% 1.7% 2.9% ⬆️ +1.2%
AWS 1.7% 2.2% +0.5%
Manager(级别) 12.3% 14.8% 16.5% ⬆️ +1.7%
Director 4.1% 4.2% +0.1%
CXO 3.2% 2.9% -0.3%

3/24 最新完整快照

数据来源:/analytics/demographic-detail/followers,2026-03-24 抓取

维度 Top 1 Top 2 Top 3 Top 4 Top 5
Job Title Search Consultant 10.1% TPM 7.7% ⬆️ Software Engineer 4% HR Specialist 2% Recruitment Specialist 1.8%
Location 杭州 19.4% Greater Seattle 8.6% ⬆️ 深圳 6.6% Greater Vancouver 6.2% ⬆️ 朝阳 5.1%
Industry Tech/Internet 32.4% HR Services 17.8% Software Dev 11.2% Staffing 8.6% IT Services 5.7%
Seniority Senior 42.3% Entry 17.6% Manager 16.5% ⬆️ Director 4.2% CXO 2.9%
Company Amazon 4.4% ⬆️ ByteDance 2.9% ⬆️ AWS 2.2% Huawei 1.3%

✅ 20 天改善总结(3/3 → 3/24)

  1. TPM 从 4.9% 升至 7.7%(+57%):同行渗透加速,差距从 2.3x 缩小到 1.3x(vs Search Consultant)
  2. Seattle 从 #3(6.6%) 升至 #2(8.6%):目标市场已超过深圳成为第二大受众区
  3. Vancouver 从 #5(5.4%) 升至 #4(6.2%):本地渗透稳步上升
  4. Seattle + Vancouver 合计 14.8%(3/3 时为 12.0%):目标区域合计份额已接近杭州(19.4%)
  5. Amazon + AWS 合计 6.6%(3/3 时为 5.1%):目标公司渗透持续上升
  6. HR/猎头合计从 29.2% 降至 26.4%:受众结构持续改善
  7. Manager 从 14.8% 升至 16.5%:管理层关注增加
  8. 杭州从 21.2% 降至 19.4%:中国受众占比持续稀释

仍需改进

  • 杭州仍然 #1(19.4%),但 Seattle+Vancouver 合计已追到 14.8%,差距缩小到 4.6%
  • Search Consultant 仍然是 Job Title #1(10.1%),但已从 12.3% 持续下降
  • Director/CXO 合计 7.1%(vs 3/3 的 7.3%),决策层渗透停滞
  • Recruitment Specialist 新进入 Top 5(1.8%),猎头换了个名字又来了

改进计划(3/24 更新)

P1 — ✅ 已完成:帖子加 Vancouver/Seattle CTA - 3/23 帖子结尾已加 "If you're a senior engineer or PM in Vancouver or Seattle..." - 加了 #Vancouver hashtag

P2 — 主动评论 Seattle/Vancouver 本地 hiring manager / recruiter 帖子(本 session 执行中) - 目标:每周 3-5 条有意义评论,被对方关注后 demographics 地理分布会自然改善 - 优先找:Amazon/AWS Vancouver/Seattle 的 TPM hiring manager

P3 — 继续发 TPM × AI 内容 - TPM 占比从 4.9%→7.7% 验证了这条路线有效 - 下一步目标:TPM 超过 Search Consultant 成为 #1 Job Title

P4 — 打入 Director/CXO 层 - 7.1% 已停滞 20 天,需要新策略 - 可能方法:评论 VP/Director 级别人物的帖子,或写"给 Director 看的 TPM 价值"类帖子


人脉拓展记录

记录所有已发送连接请求,避免重复。每次 session 后更新。

操作技巧(Playwright 自动化)

  • 连接 URL 公式https://www.linkedin.com/preload/custom-invite/?vanityName={vanityName}
    • 直接导航即可弹出 "Add a note?" 对话框,无需点击 Connect 按钮
    • 对于 Creator Mode 等隐藏 Connect 按钮的主页同样有效
  • Premium 浮层拦截:原生 click 会被 Premium upsell 遮挡 → 改用 dispatchEvent(MouseEvent) via page.evaluate() 或直接导航到 invite URL
  • 判断是否已连接:profile 主页只显示 "Message"(无 Connect)= 已是 1st degree,跳过
  • 月度个性化 note 限额:免费账号每月可附 note 次数有限 → 2026-02 已用 2 次(Lu Haibo + Bo Li),此后均 "Send without a note"
  • 搜索策略:keyword 搜索比 filter URL(company ID)效果好;geo filter 参数:geoUrn=%5B%22103644278%22%2C%22101174742%22%5D(美国+加拿大)

连接请求批次汇总

批次 日期 人数 类型 累计
批次一 2026-02-21 5 Senior TPM (Amazon/Google/Microsoft, Seattle) 5
批次二 2026-02-21 8 ex-Alibaba TPM/Tech Leader (美国+加拿大) 13
批次三 2026-02-22 6 Vancouver/Seattle TPM 19
批次四 2026-02-22 7 Vancouver/Seattle 扩展 26
批次五 2026-02-23 8 Seattle/Vancouver 扩展 34
批次六 2026-02-23 8 Seattle ex-Alibaba/TPM 42
批次七 2026-02-25 7 Vancouver/Seattle Sr TPM 49
批次八 2026-02-27 8 Vancouver/Seattle/Canada Sr TPM 57
批次九 2026-02-27 10 Vancouver/Seattle + Google/Microsoft/Amazon 67
批次十 2026-02-27 10 Seattle/US Google/Microsoft/TikTok Sr TPM 77
批次十一 2026-02-27 10 Amazon Principal + Meta TPM 87
批次十二 2026-02-27 10 Vancouver 本地科技公司 TPM 97
批次十三 2026-02-27 10 Amazon Principal + Apple + Expedia 107
批次十四 2026-02-28 7 ex-Alibaba 开发者(搜索页1) 114
批次十五 2026-02-28 5 ex-Alibaba 开发者(搜索页2) 119
批次十六 2026-02-28 5 HDU 校友(北美) 124
批次十七 2026-02-28 3 HDU 校友页2(北美) 127
批次十八 2026-03-02 1 ex-Alibaba SWE(Vancouver: Kris Yang/Fortinet) 128
批次十九 2026-03-04 20 ByteDance TPM 14 + Amazon/Vancouver/Seattle 6 148

总计:148 个连接请求(2026-03-04)


关系维护流水线

阶段 触发 动作 时间窗口 成功标记
Stage A 已发送 记录,7天内不重复触达 Day 0-7 对方接受
Stage B 对方通过 发送 follow-up(通过后 24-72h) 通过后 24-72h 收到实质回复
Stage C 有回复 分享内容或评论其帖子 回复后 3-7 天 出现第 2 次互动
Stage D 2+ 互动 进入 referral-ready 列表 M2-M6 愿意给内推

Follow-up 模板:

Thanks for connecting, [Name]. I'm mapping my transition to Senior TPM roles in Vancouver/Seattle and your path stood out. Quick take: 1) what separates strong TPM candidates at Amazon? 2) one mistake to avoid in the interview loop? Happy to share my large-scale program lessons in return.


已发送联系人跟进看板(按周更新)

姓名 公司 初次发送 是否通过 最近互动日期 当前阶段 风险 本周动作
Rose Ruoxi Liu Amazon Fresh 2026-02-25 ✅ 已通过 2026-03-05 Stage B (Phase 2) ✅ Follow-up 已发送
Jay (訾皖杰) Tze ex-Alibaba Cloud 2026-02-25 ✅ 已通过 2026-03-05 Stage B (Phase 2) ✅ Follow-up 已发送
Bing-gong Ding Microsoft 2026-02-27 ✅ 已通过 2026-03-05 Stage B (Phase 2) ✅ Follow-up 已发送
Cory Eden Spring Financial (ex-AWS) 2026-02-27 ✅ 已通过 2026-03-05 Stage B (Phase 2) ✅ Follow-up 已发送
Jagan Chebolu Amazon 2026-02-22 ❌ Pending - Stage A 12天未通过 继续等待
Meredith Underell Wrapbook 2026-02-27 ❌ Pending - Stage A 7天未通过 继续等待
Tao Song Amazon 2026-02-21 待确认 - Stage A 13天未通过 考虑归档
Lu Haibo Amazon 2026-02-21 待确认 - Stage A 13天未通过 考虑归档
Bo Li Microsoft 2026-02-21 待确认 - Stage A 13天未通过 考虑归档
Sarfaraz Sayyed Google 2026-02-21 待确认 - Stage A 13天未通过 考虑归档
Yonghua Kelly X. - 2026-02-21 待确认 - Stage A 13天未通过 考虑归档

参考文档(快速跳转)

  • 优化执行方案:03-career/amazon-pmo/linkedin_profile_optimization_20260218.md
  • Bytedance 面试备份:03-career/bytedance-pmo/
  • Amazon TPM 备份:03-career/amazon-pmo/

数据变化统计(基于基准 2026-02-20)

指标 基准值 (2026-02-20) 当前值 (2026-03-26 session 28) 变化
Profile viewers / 90d 5(7d口径) 130 —(口径不可比)
Profile appearances / week 1,138(3/10-3/17,●0%未更新)
Post impressions / 7d 1 696 ↑349.1% vs prior 7d
Members reached / 7d 309 ↑263.6% vs prior 7d
Search appearances / 7d 2 9(●0%,数据区间3/10-3/17) +7(↑350%)
Followers 356 455 +99(+27.8%
Connections 326 450 +124(+38.0%

说明:Profile views 口径从 7d 改为 90d(2026-02-21 起),无法直接与基准对比。


简短分析(2026-03-26,第 35 天 / 第 5 周)

增长结果:

从启动到现在 35 天,Followers +99(27.8%)、Connections +124(38.0%)——纯靠内容 + 人脉拓展,无付费推广。4 周目标(Followers 400+)已超额完成(455)。

内容表现 — 突破性一周:

Post impressions 7d 达到 696(↑349.1% vs prior 7d),Members reached 309(↑263.6%)—— 这是启动以来表现最好的一周。

日趋势显示 3/23 发帖日是明确拐点:Mar 20=70 → Mar 21=134 → Mar 22=176 → Mar 23=534(↑204%) → Mar 24=609 → Mar 25=683 → Mar 26=696。3/23 当天贡献了本周 77% 的增量。

关键帖子表现:

帖子 Impressions 天数 日均 特点
AI替代TPM吗 (2/27) 810 28d 28.9 最高表现帖,午夜场景+强反转
AI各工种冲击 (3/23) 306 3d 102 日均最高,首篇配图帖
16.5s→5.4s优化 (3/19) 243 7d 34.7 二次推送效应明显
AI替代讨厌的工作 (3/2) 230 24d 9.6 长尾稳定

AI各工种帖日均 102 impressions,是 810 帖日均(28.9)的 3.5 倍。如果保持当前速度,预计 5-7 天内超过 810 成为最高帖。信息图可能是核心差异因子。

Demographics 快照更新(3/3 → 3/24,20 天变化):

  • Seattle+Vancouver 合计 12.0% → 14.8%(追赶杭州 19.4%)
  • TPM 占比 4.9% → 7.7%(追赶 Search Consultant 10.1%)
  • Amazon+AWS 合计 5.1% → 6.6%
  • HR+猎头 29.2% → 26.4%(持续下降)
  • 详见 Demographics 快照对比表

Session 27-28 累计操作:

  • 发布 1 篇 LinkedIn 帖子(AI各工种冲击 + 8 roles 信息图)+ X standalone post
  • 发布 5 条评论:Alex Xu、Nick Palasz、Ajinkya More (Remitly)、Nilesh Naik、Arpit Shah (Google)
  • 更新 Demographics 快照(上次 3/3 → 本次 3/24)
  • LinkedIn playbook 新增评论规则;CLAUDE.md 同步

仍需改善:

  • Search appearances 数据区间仍停在 3/10-3/17(9次),需等下周更新
  • Profile viewers 90d = 130,增长停滞
  • Followers/Connections 日增 <1,需恢复 outbound 连接请求
  • Director/CXO 层 7.1% 未有改善

下一步:

  1. ✅ 3/23 新帖 Day 3 已达 306 impressions,日均 102(历史最高日均)—— 继续观察能否超 810
  2. ✅ Demographics 快照已更新(3/24)
  3. ✅ Vancouver/Seattle CTA 已加入帖子
  4. 下一篇帖子继续配信息图,验证配图是否是 impressions 提升的可复制因子
  5. 恢复 outbound 连接请求 —— Followers/Connections 增速已放缓到 <1/天
  6. 检查 AI各工种帖评论区是否有互动需要 reply
  7. 考虑写一篇针对 Director/CXO 的帖子("TPM 给业务带来的 ROI"),突破 7.1% 停滞

内容写作方法论

已迁移至独立文档:03-career/amazon-pmo/linkedin_writing_playbook_20260314.md


待发布内容草稿

草稿四:TPM 3问题框架 × AI Agents

状态:✅ 已发布,2026-03-16 LinkedIn: https://www.linkedin.com/feed/update/urn🇱🇮activity:7439332772328095745/ X: https://x.com/XINDR365/status/2033567645652693221

灵感来源:Stefanie Brown 的 3-questions 框架帖(108 reactions,高人气原因分析:自我诊断格式让读者对号入座,Q1反转句打中大多数PM的坏习惯,Q3有独立传播的格言句) 风向叠加:本周"Agentic Teams"爆发(多篇文章同期:Agent Teams Are Here / Agentic PM / AI coding agents debate)

LinkedIn 正文:

Managing 430 engineers on Singles' Day taught me to ask 3 questions every Friday.

Now I'm asking them about AI agents. The questions haven't changed. The answers are harder.

1. What actually changed?
Not what shipped. What behavior moved in production. Agents optimize for the objective you wrote down, not the one you meant.

2. Who knows about it?
Agents don't send status updates. They don't flag blockers. If you don't build the visibility layer yourself, no one knows what's happening — including you.

3. What breaks if we stop?
At 430 engineers, something always depended on your team without telling you. Same with agents. The dependency map doesn't write itself.

The work that used to be coordination is now instrumentation.

Same job. Different tools. Higher stakes when you miss.

What's the hardest visibility problem you've hit managing AI-driven work?

#TPM #AgenticAI #ProgramManagement

X Standalone Post(236 chars):

Managed 430 engineers for Singles' Day. Now managing AI agents.

Same 3 questions. Harder answers.

What moved? Who knows? What breaks if we stop?

Agents don't flag blockers or write status updates. The dependency map still doesn't build itself.

#TPM #AgenticAI

X 发布后在 tweet 下 Reply 放 LinkedIn URL(不放正文)


草稿二:GPT-5.4 × Agile(平台不确定性)

状态:待发布,预计 2026-03-07(已过期,待重新安排)

LinkedIn 正文:

Agile was the answer to changing requirements.

Shorter sprints. Ship every two weeks. Stop fighting uncertainty,
work with it instead.

GPT-5.4 isn't out yet. But the question it raises isn't new —
it's just landed somewhere Agile didn't plan for.

My sprint is two weeks. The model cycle is 90 days. Those don't
line up. Somewhere in that gap, the user story from sprint 1
("summarize meeting notes") hits a model upgrade that changes
the output format — and nobody has a ticket for that.

Agile handles changing requirements pretty well.

It wasn't designed for a changing platform underneath.

That's not a backlog problem. That's a planning model problem.

How are you handling AI model upgrades in your sprint cycles?

#TPM #AgenticAI #Agile

X Thread:

Tweet 1:

Agile was the answer to changing requirements.

Shorter sprints, embrace uncertainty, stop fighting the unknown.

GPT-5.4 just moved the uncertainty upstream.

Tweet 2:

My sprint is 2 weeks. The model cycle is 90 days.

Somewhere in that gap, a user story from sprint 1 hits a model
upgrade — output format changed, prompts broken — and nobody
has a ticket for it.

Tweet 3:

Agile handles changing requirements.

It wasn't designed for a changing platform.

That's a different problem.

→ [LinkedIn 链接]

#TPM #AgenticAI #Agile